The goal of this research is to implement and test an algorithm that will permit accurate recording of the Auditory Brainstem Response (ABR) at very high repetition rates, by extracting the ABR from later responses (e.g., the Middle Latency Response -- MLR). Phase I will accomplish the goal of programming the necessary stimulus and analysis algorithms, and will test the method under conditions of real noise and jitter by analyzing ABR recordings from anesthetized laboratory cats. Phase II will study the use of the method in recordings from humans, using a variety of filter settings and three or more simultaneous channels. The improved method of recording the ABR will increase the sensitivity of ABR testing in a wide variety of diseases, including HIV infection and AIDS.